1. Field of the Invention
The present invention relates to a storage structure for data on a computer, that is, a method of recording data on the memory of a computer.
2. Prior Art
As a conventional data storage structure, there is known, as shown in FIG. 8, a data structure having records/items which are predeterminedly fixed in a fixed-length binary form.
Such data structures are generally used because of certain restrictions with respect to the hardware, such as the fact that computer hardware resources (disks, memories, etc.) are expensive and hence have to be effectively utilized, as well as with respect to processing efficiency wherein data have to be processed at high speeds. The adverse effects of such known data structures are given below.
1. Data disappearance due to business programs PA0 2. Data disappearance due to update of database PA0 3. Data disappearance in data structure with fixed records/items PA0 4. Limitation of data analysis in fixed data structure PA0 5. Difficulty in system development for data structure of binary form PA0 6. Difficulty in system development for data structure of record/item PA0 7. Difficulty in system development for fixed data structure PA0 1. Storing the data in a set form of meta-data and the data, as a minimum unit. PA0 2. Where there occurs bundling, a set of data by nesting with BEGIN--END is formed for recording. PA0 3. Where not applicable to the above situations 1 or 2, the recording is by natural language (sentence).
In the computer many data occur (are input) sequentially. However, where the system has a fixed data structure as described above, data is stored at a predeterminedly fixed data structure. However, at the same time, previously unfixed data is discarded. That is, previously existing data can disappear upon preparing a business program.
Further, the record/item structures are stored by putting a plurality of items into one record. However, there are many instances where no storage is done in order of occurrence or time for each item or the like, which also results in data disappearance
Where a database is provided to carry out data analysis or various processes, the data occurring have been narrowed down to data which are required for updating the database (transaction data). In this manner, the disappearance of data has also been encountered in updating the database.
Furthermore, there are many cases in which transaction data has been deleted. If such data is left, it is rarely used for data analysis.
This is nothing other than the fact of disappearance of data, despite the fact that progress in OLTP (On-Lines Transaction Processing) has enabled updating of the database with greater reality and stability.
In many cases, the length of the record is fixed and the length of the item is also fixed. Accordingly, where unexpected content arises, data disappears or it is omitted. For example, where the commodity name item has a fixed length, such as having eighteen digits, if there occurs a commodity name with greater digits, it may be omitted. Or, where the sales results are fixed at a number of eight digits, there may be a case where the data for the number of the ninth digit exceeding the eighth digit disappears.
Data disappearing in a manner such as the above 1, 2, 3 or 4 situations is naturally impossible to analyze. If the data is to be stored, it is necessary to redesign the predeterminedly fixed data structure together with the amendment/development of programs related thereto.
Also, although data analysis is done from various points of view, the predeterminedly fixed data structure is constructed by assuming the viewpoint of analysis. To this end, if analysis is to be made from an unexpected viewpoint, the data required for such analysis has to be prepared. If not prepared, the analysis is impossible carry out.
The binary-formed data structure is impossible to read at one view and can be comprehended only by those people with expertise, thus largely limiting the available personnel for development. Also, to save resources in respect of hardware, efforts are made to shorten the record length and the item length as much as possible so that the system design is made such that reference to the master file or the like is done using many codes and flags. To this end, there arise additional processes between a plurality of files, which further complicates the system and makes development difficult.
The contents of data cannot be comprehended and dealt therewith, unless there is present a person who knows the design chart (file/layout diagram) or the design chart itself. This limits the available development personnel to those people who know the design of the record/item.
The fixed data structure has a data structure in which results thereof are previously assumed. The entire system is also designed like this. As a result, they have only minimal flexibility for unexpected items or additional processes. In worst cases, a redesign of the entire system is required. For example, only increasing the number by one digit for items, such as commodity codes, of highly frequent use requires the amendment of related systems such as on-screen input, display, printing, etc.
The present invention has been made in view of the above points, and it is the object of the invention to provide a novel data structure which is capable, without disappearance, of storing in operation or process unit form all the data occurring (input) on a computer system in its order of occurrence and with such a flexible structure that it can be comprehended by anyone at a look, thus enabling easy alteration of the system as required.